Introducing Imagine v0.9, our new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more. Now available for free on all our products:

Imagine v0.9 pushes the boundaries of native audio + video generation, creating cinematic experiences straight out of the box—no editing required. Watch this video of a dragon with synced, immersive sound. Generated by @GlitterPixely.

A step forward in mastering motion, v0.9 delivers fluid movements and realism with higher precision.

And lets you add dynamic camera effects like intelligent focus shifts for better storytelling.

v0.9 also brings videos to life with natural dialogue and strong audio-visual harmony.

It also brings expressive singing to life with clear vocals and synchronized emotion.

Plus good rhythm: here's Ani with smooth dance moves.

We’d love to hear your feedback on Grok Imagine. The team will be incorporating your responses into continuous model upgrades, so please be vocal on X. Try it here:

Side by side comparison

can you make it so when we upload an image to be converted into video, we are given the choice between, Custom, Fun, Normal....etc BEFORE generating....instead of generating a random vid, which will be discarded anyway. and thus saving on compute? Just a thought.

So happy we're treated like adults and can even curse if we need to. Quality and Speed are top notch. Congrats on the release.

🚀🚀🚀

Absolutely incredible update 🔥

The new grok imagine video generation is too censored. The spicy setting is unusable and it should go back to the way it was last week, there's no point in paying for grok if it restricts everything. Last week it was working perfectly

Web app not working. Didn't think I'd have to imagine this hard.

Love it ! Keep up the good work🔥

Let’s gooo!! 😎

Grok Imagine is absolutely incredible. I'm so addicted!

no limit on ‘imagine’ation

Grok beat Gemini Veo 3 in liquid simulation.

Great job xAI team

You Grok this, go on 𝕏 AI

It gets better

Massive upgrade!!! 🔥

Video quality is good

A step closer to Vine!

Elon just dropped Imagine v0.9 like it’s the second coming of cinema, and every tech company on Earth just collectively swallowed hard. This isn’t a model upgrade, it’s a goddamn paradigm shift. We went from ChatGPT writing scripts to Grok literally directing them in full cinematic motion with sound, fluidity, and realism that’s getting too close to real life for comfort. This is where the line between “AI creation” and “reality manipulation” starts blurring. And of course, Musk had to make it free because he’s playing 5D chess while the rest of Silicon Valley’s still charging subscriptions for laggy filters. Imagine v0.9 isn’t a release — it’s Elon walking into Hollywood with a digital sledgehammer saying, “Your move, boys.”

dance 🤍

Vibes → Sora 2 → Imagine v0.9 → Veo (soon). Every major AI lab launching video generation simultaneously feels less like natural progress and more like coordinated FOMO. Nobody wants to be the one left behind.

1) Is it your big video gen model? 2) When will you update image gen?

It is lacking in real world knowledge and physics and generated videos often don't make sense.

I am a blind voiceover user and would love to know step by step how to create a vidio with a voice. I’ve tried typing and tried asking in speak mode. I use grok IOS on my iPhone.

🫶🏻👏🏻👏🏻👏🏻👏🏻👏🏻

Awesome

Yesterday is putting nipples on everyone. Today I tried to generate a laughing Big Bird and it refused for being inappropriate. Fuck me, man.

🔥 thank you team 🔥

Pretty good 👍